ACI STRUCTURAL JOURNAL
Title no. 91-816
Seismic Load Tests on Reinforced Concrete Columns
Strengthened by Jacketing
by M. Rodriguez and R. Park
Four reinforced concrete column units were tested subjected to simulated
seismic loading to investigate repair and strengthening techniques. The as-
built columns were 350 mm ( 13.8 in.) square and contained low quantities
of transverse reinforcement as was typical ofbuilding columns designed and
constructed prior to 1970. The column units represented the column region
between the midheights of successive stories. A stub was present at the mid-
height of each unit to represent a portion of the two-way beams and slab at
the beam-column joint. Two column units were tested, repaired, and strength-
ened by jacketing and retested. The other two column units were strength-
ened by jacketing and tested. The jacketing consisted ofa 100-mm (3.94-in.)
thickness of added reinforced concrete. The new longitudinal reinforcement
was placed through the floor slab. Two arrangements of transverse rein-
forcement in the jacket were investigated. The as-built columns displayed low
available ductility and significant degradation of strength during testing,
whereas the jacketed columns behaved in a ductile manner with higher
strength and much reduced strength degradation. The retrofit of columns
using reinforced concrete jackets was found to be successful but labor-in-
tensive.
Keywords: columns (supports); concretes; ductility; earthquake-resistant struc-
tures; flexural strength; loads (forces); reinforced concrete.
Seismic design procedures have advanced considerably
since about 1970. The main developments have been in the
understanding of the nonlinear dynamic response of struc-
tures, the introduction of capacity design procedures, and the
methods for detailing reinforcement in concrete structures to
achieve the ductile behavior necessary to survive severe earth-
quakes.
Retrofitting of structures has been undertaken in several
earthquake-prone countries after structural damage caused by
strong earthquakes 1-3 or because existing structures were re-
quired to comply with more recent seismic code provisions. 4 ·5
Deficienices often found in typical existing moment-resisting
frames are inadequate shear strength of beams, columns, and
beam-column joints, and inadequate flexural strength and
ductility of columns.6· 7 A major deficiency has been inade-
quate quantities and/or anchorage of transverse reinforcement
in members and joints, particularly in columns.
Several techniques for the repair and strengthening of struc-
tural elements, such as reinforced concrete columns, have
been suggested in the literature.t,s-to One retrofit technique
150
TECHNICAL PAPER
for buildings has involved the jacketing of columns with ad-
ditional (new) reinforced concrete. This retrofit procedure
was widely used in Mexico City after the 1985 earthquake. 10
The use of column jacketing has the advantage that the re-
sulting increase in the lateral load resistance of the building
is distributed throughout the structure, and therefore that new
foundations, or significant strengthening of existing founda-
tions, may be avoided. to
This paper reports the results of an experimental study of
the improvement in seismic behavior of reinforced concrete
columns repaired and/or strengthened by concrete jacketing.
The as-built columns tested were typical of those constructed
prior to the 1970s. The tests involved both the as-built
columns and the columns strengthened by concrete jacketing
with added longitudinal and transverse reinforcement.
RESEARCH SIGNIFICANCE
This research work forms part of a research program aimed
at developing methods for retrofitting existing reinforced con-
crete structures to increase their seismic resistance where nec-
essary. Such research is required because there is a need for
more experimental and analytical investigations to provide
designers with further information regarding the seismic be-
havior of structures repaired and strengthened by different
retrofitting techniques. It is evident that many of the tech-
niques that have been used in earthquake-prone countries
have been based mainly on engineering judgment. 4-5 In par-
ticular, although the jacketing of columns with new reinforced
concrete has been commonly used in some countries, there
has been limited experimental testing of the effectiveness of
such jacketing. The aim of this research is to investigate the
increase of strength, stiffness, and ductility, which can be
achieved by jacketing existing damaged or undamaged rein-
forced concrete columns with new reinforced concrete.
ACI Structural Journal, V. 91, No.2, March-Aprill994.
Received Dec. 21, 1992, and reviewed under Institute publication policies. Copyright ©
1994, American Concrete Institute. All rights reserved, including the making of copies un-
less permission is obtained from the copyright proprietors. Pertinent discussion will be pub-
lished in the January-February 1995 ACI Structural Journal if received by Sept. I, 1994.
ACI Structural Journal I March-April 1994
Mario Rodriguez is a professor in the lnstituto de lngenieria, Universidad Nacional Au-
tonoma de Mexico at Mexico City. He has teaching and research interests in reinforced
concrete structures, particularly in earthquake-resistant design. He is a member of ACI 1
10'-0" to
Committees 341, Earthquake-Resistant Concrete Bridges; 368 Earthquake-Resisting Con-
next level
crete Structural Elements and Systems; and 369, Seismic Repair and Rehabilitation. During
the period 1989-1990, he was on leave at the University of Canterbury, New Zealand.

Robert Park, FACI, is Deputy Vice-Chancellor and a professor of civil engineering at the
University of Canterbury, Christchurch, New Zealand. He has teaching and research in-
terests in reinforced and prestressed concrete structures, particularly in design for earth- 1~dia. bars
quake resistance. He is a member of ACI Committees 341, Earthquake-Resistant Concrete ~ '4jia. ties

~16"x 16"Column
Bridges; 352, Joints and Connections in Monolithic Concrete Structures; and 368, Earth-
quake-Resisting Concrete Structural Elements and Systems; and joint ACI-ASCE Committee
441, Reinforced Concrete Columns. Professor Park is the author or coauthor of several S£CnON u-u
technical papers and two books. He has been a corecipient of several awards, including
ACI's 1984 and 1989 Raymond C. Reese Research Awards.
12'

DETAILS OF COLUMN TEST UNITS 12"crs

Prototype column
A typical reinforced concrete column from the lower story 1": 25.t.mm
1' : 0.305m
of a seven-story moment-resisting frame constructed in New
Zealand in the late 1950s is shown in Fig. 1. The columns in
some regions of this building have inadequate flexural
strength to prevent plastic hinging of columns from occur-
ring during a severe earthquake. Also, the transverse rein-
forcement in the columns is inadequate according to current Fig. ]-Dimensions and reinforcement details for typical column
seismic codes. Although built in New Zealand, the column of a moment-resisting frame designed in the late 1950s
shown in Fig. 1 is typical of many in other countries before
current code provisions for seismic design were introduced. 60
350
5
As-built Units 51 through 54 hoops!i~
R6265crs
at
~!"'
Four reinforced concrete column units, referred to as Units Long. ~ ~
S1, S2, S3, and S4, were constructed at a 7/8th scale to rep-
8-R20~
Remf. 8-R20
resent the as-built prototype column shown in Fig. 1. The SS
39mm
cross sections of the column units were 350 mm (13.8 in.) --:;, cover SECTION
A A-A
square. Each column unit had a height of 3.3 m (10.8 ft) and
represented the column region between the midheights of suc-
cessive stories of the frame. Typical dimensions and rein- 3600
forcing details of the as-built column units are shown in Fig.2. 450
Eight longitudinal bars were distributed evenly around the
perimeter of the column cross section, and the transverse re-
inforcement consisted of sets of overlapping square hoops, as
shown in Fig. 2. The stub at the midheight of the column units
represented a small portion of the two-way beams and slab at 1425
the beam-columnjoint. The longitudinal and transverse rein- Dimensions ore in mm
forcement of the column units were plain round bars, as was Tmm: 0.0394 in
typical of the 1950s, of Grade 300 (jy ~ 44 ksi) steel. The con-
crete of column units was normal weight with a design con- 150 •~o.Z:::~H--- Sleeve for
....1....;=-L._____ SOmm pin
crete compressive strength// of 20 MPa (2900 psi). The con-
10mm Anchor
crete was cast with the columns in the horizontal position. Plate
Table 1 lists details of the reinforcement for the as-built
Units S 1 through S4. Table 2 lists details of concrete com- Fig. 2-Dimensions and reinforcement details for the as-built
pressive cylinder strengths and the axial load ratios applied column units
to columns during testing. Table 1 shows that the quantities
of transverse reinforcement present in the as-built column Jacketed Units SS1 through SS4
units were very low compared with the quantities required by The as-built Units Sl and S4 were first damaged, by sub-
the ACI Building Code 11 for confining the concrete and pro- jecting them to simulated seismic loading, and then jacketed
viding lateral support to the longitudinal reinforcement. How- to become Units SS 1 and SS4, respectively. The as-built Units
ever, the quantities of transverse reinforcement were adequate S2 and S3 were jacketed without first testing to become Units
for shear resistance. The theoretical nominal shear strength SS2 and SS3, respectively.
of the as-built column units computed using the ACI code! I Units SS1 and SS2 had 100-mm (3.94-in.)-thick concrete
approach was 1.3 times the shear force required to develop jackets containing eight new longitudinal bars bundled into
the theoretical nominal flexural strength. the corners of the jacket and new square hoops, as shown in
ACI Structural Journal I March-April 1994 151
-
Table 1 - Details of reinforcement in column test units
Transverse reinforcement
Longitudinal reinforcement in potential plastic hinge regions

Ash
db, Jy, pr, db, hh, Sh, Shh
11
, Ash
Unit Part of column mm MPa percent mm MPa mm percent Ash, ACI

S1,S2,S3,S4 As-built column 20 325 2.05 6 350 265 0.13 0.10

SS1 Column jacket* 16 502 1.36 10 340 95 0.35 0.38
SS2 Column jacket* 16 502 1.36 10 340 95 0.35 0.40
SS3 Column jacket* 12 491 1.28 10 330 72 0.82 1.00
SS4 Column jacket* 12 491 1.28 10 330 72 0.82 0.79
*Units SS1 and SS4 were repaired and strengthened; Units SS2 and SS3 were strengthened.
Note 1 rom= 0.0394 in., 1 MPa = 145 psi.

Table 2 - Compressive strength of concrete in
column test units
Concrete Compressive
cylinder strength at stage of testing
Strength
Unit Age, days /c~ MPa
S1 110 29.5
S4 104 25.9
SS1 152* 32.9*
SS2 75* 34.0*
SS3 77* 19.4*
SS4 30* 25.2*
*Data for the concrete jacket of the column.
Note: 1 MPa = 145 psi.
Fig. 3(a). Units SS3 and SS4 had 100-mm (3.94-in.)-thick
concrete jackets containing 12 new longitudinal bars distrib-
uted around the perimeter of the cross section of the jacket
and new sets of overlapping square and octagonal hoops, as
shown in Fig. 3(b ).
The longitudinal reinforcement in the jackets were de-
formed bars from Grade 430 ({y~ 62 ksi) steel, and the trans-
verse reinforcement were plain round bars from Grade 300
({y ~ 44 ksi) steel. The concrete used for the jackets of these
column units was normal weight and was cast with the
columns in the vertical position. Before placing the concrete
jackets, the surface of the concrete of the as-built coluqms
was lightly roughened to an amplitude of about 2 to 3 mm
(0.08 to 0.12 in.) by chipping, and in the case of the previ-
ously damaged columns of Units S1 and S4 all loose concrete
was removed. Fig. 4(a), (b), and (c) show, for Unit S1, the
damaged region of the as-built column above the slab after
the initial seismic load testing, the damaged region of the as-
built column after removing the loose concrete, and the
column with the new reinforcement before placing the con-
crete jacket, respectively. The final placement of concrete for
the jacket around the column below the floor slab was through
holes made in the slab through which the new longitudinal
column bars passed. Care was needed to insure that the con-
crete was adequately compacted and that the concrete in the
jacket reached the underside of the slab.
Tables 1 and 2 list for Units SS 1 through SS4 the details
of the reinforcement and the concrete compressive cylinder
strengths. The axial load ratios applied during the tests are
also shown. The design of longitudinal and transverse rein-
forcement in the concrete jackets complied with the require-
152
ments for ductile columns designed by capacity design for
seismic loading according to the New Zealand concrete de-
sign code NZS 3101,12 except that in the columns of Units
SS 1 and SS2 the horizontal spacing of the tied longitudinal
_f__ bars exceeded the code-permitted maximum spacing of 200
fc' Ag
mm (7.9 in.).
0.2 Table 1 shows that the quantities of transverse reinforce-
0.2 ment in the jackets of Units SS 1 through SS4 were generally
0.1 less than that required by the ACI code 11 for confining the
0.1 concrete and providing lateral support to the longitudinal bars.
However, the quantities of transverse reinforcement in the
0.1
jackets were adequate for shear resistance. The ratios of the
0.1 theoretical nominal shear strength of the jacketed column
units computed using the ACI codell approach to the shear
force required to develop the theoretical nominal flexural
strength were about 1.5 for Units SS1 and SS2 and about 2.7
for Units SS3 and SS4.
In the beam-column joint regions, the longitudinal column
bars in the jackets were laterally restrained by ties, which in
the case of Units SS 1 and SS2 were welded to bolts anchored
in the concrete of the as-built units [see Fig. 4(d)]. In Units
SS3 and SS4 the lateral restraint to the longitudinal column
bars in the beam-column joint regions was applied more pos-
itively by hoops that were made up by bars passed through
holes drilled horizontally through the concrete of the beams
and welded in place to form hoops.
TESTING OF THE COLUMN UNITS
Simulated seismic loading
The column units were subjected to simulated seismic
loading. The method of loading is shown in Fig. 5. Quasi-
static cyclic lateral loading H was applied to the stub at the
midheight of the unit through a loading frame. A universal
testing machine was used to apply a constant axial compres-
sive load P through steel rollers and end plates-to each end of
the column. For the as-built columns, the axial compression
load ratio P/A 8/c' was equal to 0.2, which was typical of the
lower story columns of the building investigated. For the re-
paired and/or strengthened units, this ratio was equal to about
0.1 averaged over the jacketed column (Table 2). The columns
were in an unloaded condition while the jackets were being
constructed. Note that in an actual structure the original
column would continue to carry gravity loading and the jack-
eted portion would ony assist in carrying forces applied due
to lateral loads, unless the floors were jacked apart and the
gravity load was carried by props while the columns were
ACI Structural Journal I March-April 1994
being jacketed.
As is evident from Fig. 5, the lateral loading applied to the
Reinforcement of RIO ties welded Dimensions ore in mm
column units caused single curvature bending over the whole stub shown in to anchored bolls
tmm=D.0394in
in the beam concrete
height of the unit for a particular direction of lateral loading.

~INAl
This differs from the actual case in a moment-resisting frame
where, during lateral loading, the input moments of the beams
causes the sign of the curvature to be different above and lUJ ~~~MN STUB
below the beam-column joint. In tests investigating the be-
havior of beam-column joint cores, a method of loading that
causes a.change in the sign of the column curvature above and
1650
below the joint is necessary, because of the significant shear
and bond stresses induced by the change in the sign of the cur-
vature. However, in tests investigating the behavior of the
plastic hinge regions at the ends of columns, the simulation
of the exact stress conditions in the beam-column joint core
is not so important. The loading arrangement shown in Fig.
5 has often been used in the past to investigate columns.I3,14
In all tests, a cycle of lateral loading to 0.75HAci was ini-
tially applied, where HAci is the calculated lateral load asso- fa} IUNITS SS 1 & SS2j
ciated with the nominal theoretical flexural strength Mi being
reached at the critical sections of the column, computed using
the ACJII rectangular compressive stress block for the con-
crete with an extreme fiber concrete compressive strain of 550
1
0.003, the measured concrete compressive cylinder strength, Reinforcement of
RIO Hoops placed
th rough holes
stub shown in
a strength reduction factor <1> of unity, and the measured stress- original column
dril led in the
con crete
un1ts
strain relationship for the longitudinal reinforcement. Note LOADING
that the New Zealand concrete design code12 uses the same
--
DIRECTION
.. ~
-:t
assumptions for flexural strength calculations as the ACI
code 1L
5L /~ 4spaces
a 72
___1_40

-r
i2
Transverse
The yield displacement ~Y for the retrofitted Units SSl, Reinf RIO I 6 spaces
SS2, SS3, and SS4 was found from the stiffness at a lateral
....... a
displacement measured at the central stub at 0.75HAci or first
yielding of the longitudinal reinforcement, whichever was 1650 Longitud;;;;;
Reinf HD/2
5spaces
less, extrapolated linearly to HAcJ. This definition of the yield a 145
displacement ~Y is illustrated in Fig. 6 and has been widely
\ ~
used in New Zealand in the experimental testing of structural Sleeve for
50mm pin...._ )( 40
assemblages. 13,1 4 The yield displacement ~Y defined in this I'-- ~) 0
HD ,_
manner is, in fact, the displacement at first yield of an equiv- 150. t-..." 08
COLUMN SECTION
alent elastoplastic system with reduced elastic stiffness to take I 0 mm An cho r ) Dimensions ore in mm
Plate t mm =0.0394 in
into account the effect of cracking and other nonlinear elastic
effects. (b} IUNITS SS3 & SS4l
Nevertheless, because of the significant softening of the
lateral load-versus-displacement relationship measured
Fig. 3-Dimensions and reinforcement details for Units SSJ
during the tests on the as-built Units S 1 and S4, caused mainly through SS4
by inadequate bond, ~Y in these units was defined as the cen-
tral stub displacement measured at 0.75HAcJ. lengths immediately above and below the column stub were
The applied cyclic loading in the inelastic range was dis- seated directly against the faces of the stub, and, hence, these
placement -controlled. The column units were subjected to readings included deformations of the column due to bond
two loading cycles to each of Jln = ±1, ±2, ±3, ±4, etc., where slip of the longitudinal reinforcement in the stub and yield
Jln is the nominal displacement ductility factor defined as penetration of this reinforcement into the stub.
M ~y, where ~ is the lateral displacement of the central stub. Also, electrical resistance strain gages were attached in
pairs at various locations on the hoops and on the longitu-
dinal reinforcement in the plastic hinge regions.
Instrumentation
The horizontal displacement and the rotation of the central PERFORMANCE OF THE AS-BUILT COLUMN
stub were measured by a set of three linear potentiometers. UNITS S1 AND S4
Additionally, 12 pairs of potentiometers were used to mea- Load-versus-displacement response
sure the average section curvatures over 80 and 160-mm (3.15 Fig. 7 shows the experimental lateral load-versus-lateral
and 6.3-in.) gage lengths in the plastic hinge regions adjacent displacement hysteresis loops measured for. Units S 1 and S4
to the central stub. The pairs of potentiometers in the gage representing the as-built column. Also shown is the nominal
1
ACI Structural Journal I March-April 1994 153
 (a) Damaged Region of the As-Built
Column After Testing
(c) New Reinforcement Placed Before
Casting the Concrete Jacket
Fig. 4-Initial damage and method of repair and strengthening of the as-built Unit SI
ideal theoretical ultimate lateral load HAci calculated using
the ACI codell approach previously described. This theoret-
ical load is plotted as dashed lines that reduce with increase
in displacement due to the P-d effect. Fig. 7 shows that un-
like well-confined columns, where significant flexural over-
strengths have been measured 14 mainly due to increased con-
crete compressive strength and ductility due to confinement,
the measured maximum moments of Units S1 and S4 were
almost equal to the nominal theoretical strengths calculated
including the P-fl effect.
Fig. 7 shows that Units S 1 and S4 demonstrated a signifi-
cant reduction of strength after they reached the measured
maximum moments which ocurred at 1-ln equal to approxi-
mately 3. This was the stage of significant crushing of cover
and core concrete. The damage was concentrated mainly in
154
(b) Damaged Region of the As-Built
Column After Removing Loose
Concrete
(d) Method of Attaching New Ties
in the Beam-Column Joint
Region
the plastic hinge region above the central stub in the case of
UnitS 1 [see Fig. 4(a)] and in the plastic hinge regions above
and below the central stub in the case of Unit S4.
Fig. 7 also shows values for the real displacement ductility
factor 1-lr, which includes the effect of the rotation of the cen-
tral stub due to the plastic hinge deformations concentrating
either above or below the stub. As is shown in Fig. 8, 1-lr can
be calculated as 1-lr =(fl.+ 8£)/fl.y where e is the measured ro-
tation of the stub and eis the distance from the center of the
stub to the pin at the end of the column. As Fig. 7 shows, the
value for 1-ln of 3 for Units S 1 and S4 corresponds to values
of 1-lr that ranged from 3.6 to 3.9. It is considered that the
values of 1-lr attained by the columns give a better indication
of the available ductility, since 1-lr is a measure of the ductility
available from the plastic hinge undergoing the greatest
ACI Structural Journal I March-April 1994
 p REAL DISPLACEMENT DUCTILITY

~H~~-------f
FACTOR (IJ.rl. (Top side. second cycle)
1 2.1 3.9 6.4
I I I I
3 4 =lln
~ _lj_

Pin ~300
:X:
I I Ideal strength
/including
c:i
1425 ~ 200
--l_p.!J effect

-H_l 450 3600 -40

INTERS TOREY
DRIFT
3X

1 1425 /"'---..:---

~H l~~sions
Ideal strength
including I I I
P-!J effect 1 -300
IJ.n=-4 -3 -2 -1
1 I I I I 1mm =0.0394in
-5.8 -3.9 -2.2 -1 1kN = 2241b
REAL DISPLACEMENT DUCTILITY
FACTOR (IJ.r). (Top side.second cycle)

in mm fa J IUNIT S 1j
350 1mm =0.0394in

Fig. 5-Method of loading the column units

REAL DISPLACEMENT DUCTILITY
LATERAL FACTOR (IJ.r). (Bottom side.second cycle)
1 2.2 3.8 5.9
LOAD I I I I
2 3 4 = lln
Nominal strength HAC!
I I I Ideal
Strength
/including
First yielding or 0.75~ct .:::.;--::::::=-or----P-!J effect
whichever is less

-3X
~40

DISPLACEMENT.!::. (mm)

, _ ______
Ideal strength
including
P-!J effect
I I I I
First yielding or 0.75HAct lln=-4 -3 -2 -1 1mm =0.0394 in
whichever is less I I I I 1kN =2241b
-5.3 -3.6 -2.2 -1
REAL DISPLACEMENT DUCTILITY
Nominal strength HAct FACTOR (IJ.r}.( Top side. second cycle)

(b) IUNIT S41

Fig. 6-Definition of yield displacement13J4
plastic rotation, whereas Jln is a measure of the average duc-
tility available from the two plastic hinges of the column unit.
A measure of the available ductility of the column units
during cyclic lateral loading is the available displacement
ductility factor J..la, defined 13 for four loading cycles (that is,
eight loading runs) as
(1)
where LJl is the cumulative displacement ductility factor for
loading runs in which the lateral load did not reduce to less
than 80 percent of the maximum applied lateral load. Using
this definition, for UnitS 1, there were four loading runs to Jln
= 2 and four loading runs to Jln =3, giving Jla =(4 x 2 + 4 x
3)/8 = 2.5. For Unit S4, there were four loading runs to Jln =
ACI Structural Journal I March-April 1994
Fig. 7-Measured lateral load-versus-lateral displacement hys-
teresis loops for the as-built column units
2 and three loading runs to Jln = 3, giving Jla = (4 x 2 + 3 x
3)/8 =2.1.
The New Zealand concrete design code NZS 310112 spec-
ifies that structures with adequate ductility should reach a lat-
eral displacement of at least 4 to 6 times the displacement at
first yield during four loading cycles, without significant re-
duction in strength. It is evident that the measured available
displacement ductility factors for Units S 1 and S4 fell well
short of those code specified values for ductile structures.
Also, the measured ductilities for Units S 1 and S4 fell short
of those measured in previous research projects conducted at
the University of Canterbury 14 -15 during qQasi-static cyclic
loading tests on reinforced concrete columns having low axial
155

Fig. 8-Effect of unsymmetrical plastic hinge rotation on the dis-
placements of the column units
loads and containing greater quantities of transverse confining
reinforcement.
The maximum interstory drift reached by Units Sl and S4
before the lateral load reduced to less than 80 percent of the
maximum applied lateral load was about 1.9 percent. The in-
terstory drift is obtained by dividing the interstory horizontal
displacement by the story height. Interstory drift has been
suggested as a suitable index for the level of deformation im-
posed on test structures or structural subassemblages.I6 How-
ever, caution must be adopted in the use of this index in tests
because the imposed deformation needs to be related to the
stiffness of the structure and the displacement ductility
factor. 13 Noting that Units S 1 and S4 were relatively flexible,
the interstory drift of 1.9 percent attained by these two units
before substantial strength degradation is relatively small
compared with the values of at least 2 to 3 percent obtained
in previous tests on stiffer reinforced concrete columns de-
signed according to the New Zealand concrete design code12
conducted at the University of Canterbury.I4,15,17
Measured strains and curvatures
For each of Units S 1 and S4, the longitudinal strains in the
concrete core of the columns in the potential plastic hinge re-
gions were calculated from the potentiometer readings and
compared with the longitudinal strains measured on the rein-
forcing bars by electrical resistance strain gages. The com-
parisons are shown elsewhere. 18 Large differences between
these concrete and steel strains were observed, apparently due
to bond degradation occurring between the concrete and the
plain round longitudinal bars during the tests.
Evidence of bond degradation was also given by the cur-
vatures calculated from potentiometer readings. 18 Most of the
inelastic curvature in each plastic hinge region was concen-
156
trated over a relatively small length of column of about 0.5hc,
where he is the column depth, which was evidently due to
bond degradation leading to one or two main cracks.
The transverse strains measured on the hoop sets in the
plastic hinge regions of Units S 1 and S4 seldom reached the
yield strain, even in the final stages of testing.
PERFORMANCE OF THE REPAIRED AND/OR
STRENGTHENED COLUMN UNITS
Load-versus-displacement response
Fig. 9 shows the experimentallateralload-versus-lateral
displacement hysteresis loops for the retrofitted Units SS 1
through SS4. The measured lateral load-versus-displacement
hysteresis loops for the four jacketed columns indicated good
energy dissipation and only a little reduction in strength up
to the end of testing. The increase in stiffness, strength, and
ductility of the jacketed columns can be observed by com-
paring Fig. 7 and 9. The comparison indicates, for example,
that the strength and stiffness of the jacketed Unit SS 1 were
about three times those for the as-built UnitS 1. The ductility
achieved by the retrofitted columns was at a level satisfactory
for ductile structures. The maximum interstory drift reached
was about 2.8 percent.
The experimental loops shown in Fig. 9(a) and (d) for Units
SS 1 and SS4, which were damaged before retrofitting, are
quite similar to those for Units SS2 and SS3, which were not
damaged before retrofitting. This suggests that the previous
damage to the as-built columns of Units SS 1 and SS4 did not
significantly influence the performance of the retrofitted units.
Also, the experimental loops shown in Fig. 9 for Units SS 1
and SS2, which had longitudinal bars bundled into the cor-
ners of the jackets [see Fig. 3(a)], and for Units SS3 and SS4,
which had longitudinal bars distributed around the perimeter
of the jacket [see Fig. 3(b)], were not significantly different.
Hence, the bundling of bars in the comers of the jackets of
Units SS 1 and SS2 had no detrimental effect on the seismic
performance of those units, in spite of the fact that these bun-
dles were 438 mm (17.2 in.) apart and, hence, exceeded the
permitted New Zealand codel2 maximum spacing between
tied column bars of 200 mm (7 .9 in.). Similar results for bun-
dled column bars have also been found in cyclic lateral
loading tests of jacketed reinforced concrete columns con-
ducted at the University of Texas.l6
In addition, these tests indicated that the greater quantity
of transverse reinforcement in the jackets of Units SS3 and
SS4 than in the jackets of Units SS1 and SS2 (see Table 1 and
Fig. 3) resulted in no significant improvement in the seismic
behavior of Units SS3 and SS4. That is, for these jacketed
columns with relatively light axial loads of 0.1/c'A 8 , the lat-
eral load tests showed that the quantity of confining rein-
forcement recommended by the ACI code 11 is unnecessarily
high. The conservative nature of the ACI-recommended quan-
tity of confining reinforcement for columns with small axial
load levels has been discussed previously.l9 The amount of
confining reinforcement required for columns with relatively
light axial loads of 0.1/c'A 8 according to the New Zealand
code 12 can be up to 40 percent less than that required by the
ACI code. 11
In the test on Unit SS 1, the bundled ~ongitudinal bars in the
column above the central stub buckled at a nominal dis-
ACI Structural Journal I March-April 1994
 REAL DISPLACEMENT DUCTILITY
FACTOR (1.1.,), (Bottom side. second cycle)
I 2.6 4.4 6.6 8.7 10.2
.... I I I I I I
~ H1 2 3 4 5 6 : I.Ln
;; ~ I I I I I (:deal strength
~- 600 1.0- _ _ _ ..L!.~~~u:k~ft
Fracture of
longitudinal bars

-4% {'-....Fracture of
Long. bars
-3:1. 3 :;. INTEff#f1fEY
~-~~~~~~~~~~--'40r-~

I I I I
i.J.,=-6 -5 -4 -3 -2 -1 /.L
-5~5 -5.'8 -416 -3~6 -2.15 - J o.
lmm = 0394 in Ideal strength
REAL DISPLACEMENT DUCTILITY
FACTOR fi.J.r). (Bottom side, second cycle/
2
lkN = 2 4lb including P1 eftt I I I I
1mm :0.0394in
l-ln=-6 -5 -4 -3 -2 -1
fa} IUNIT SS1\ I I I I I I lkN : 224/b
-11.0 -8.9-6.8-4.8 -3.1 -1
REAL DISPLACEMENT DUCTILITY
FACTOR fi.LrJ. (Bottom side.second cycle)

REAL DISPLACEMENT OUCTILITY

FACTOR I 1-lri. !Bottom side. second cycle I
I 2.4 4.1 6.0 9.0 12.4
(C) IUNIT SS3I
I I I I I I

~ 1000
_:n'i f I i i i REAL DISPLACEMENT DUCTILITY
FACTOR ( I.J.r), (Bottom side. second cycle)
~ 800 H1.'i)__:;.r.v--.1-~
c;; I 2.8 4.3 5.8 7.9 9.9
8 500 I I I I I I
~
c:i 600
~Jc,(
1.0_
f I i f r= 1.1. n

·3%
-40
s - ~eo/ strength
including
P-!J effect
INTERS TOREY
DRIFT
-3:1. 3:1.
;r;::~;;;)._ =---rco:,...f'.,-.;.Q?,.J. -40
inclvding
Po/Jeffect
Fracture of
1
!Jn~-5 -4 -3 -2 longitudinal bars
I I 1 1 tmm ~o.0394in
-5.6 -4.6 -3.5 -2.2 I lkN ' 224/b
REAL DISPLACEMENT OUCT/LITY
FACTOR 11-1,1. IBo"om side. seco,.....nd_cy:.....ct-'el _ _...,

(b) IUNIT SS2l

I I I -3I -2I -1I
I.J.n:-6 -5 -4 lmm: 0.0394in
I I I I I I 1kN :224/b
-10.0-7.6-5.7-4.1 -2.4 -I
Fig. 9-Measured latera/load-versus-lateral displacement hys- REAL DISPLACEMENT DUCTILITY
teresis loops for the jacketed column units FACTORfi.LrJ. (Bottom s~id;:..:e·.:;;sec=on.:.;:;d..::L::.~---.

placement ductility factor Jl.n of about 6 and eventually frac-

Fig. 9-Measured lateral load-versus-lateral displacement hys-
tured [see Fig. 9(a) and 10(a)]. Although most of the damage teresis loops for the jacketed column units
was concentrated above the central stub of this unit, some
damage was also observed below the stub.
The test on Unit SS2 was terminated when a significant re- Measured strains and curvatures
duction of lateral load capacity occurred as a result of buck- The curvatures obtained from potentiometer readings and
ling of the bundled longitudinal bars in the beam-column joint the curvatures predicted by the cyclic moment-curvature
region at a Jl.n of about 6 [see Fig. 9(b)]. These bars lost their theory are shown compared elsewhere. 18 Again, there was ev-
restraint against buckling when the anchor bolts, to which the idence of slip of the plain round bars in the column.
ties in the jacket in the beam-column joint were welded, The measured steel strains on the instrumented hoop sets
pulled out of the concrete of the as-built unit. The more reli- in the jackets of the plastic hinge regions of the four retrofitted
able lateral restraint obtained from hoops, made up by passing columns indicated that these strains seldom reached the yield
bars through horizontal holes drilled in the beams of the strain, even in the final stages of testing.
beam-column joint region, and welded in place to form hoops,
used for Units SS3 and SS4, is therefore preferred. CONCLUSIONS
The damage to Units SS3 and SS4 was concentrated in the 1. The seismic load tests on two column units, representing
columns below the beam-column joints and the longitudinal reinforced concrete columns designed and constructed in New
bars buckled and eventually some fractured at Jl.n values of Zealand in the 1950s, showed that columns designed to early
about 6 [see Fig. 10(b)]. seismic codes could have low available ductility. The columns
ACI Structural Journal I March-April 1994 t57
 (a) Unit SSl
Fig. 10-Damage of the repaired and strengthened column units at the end of testing
tested were 350 mm (13.8 in.) square, contained plain round
longitudinal bars, and had transverse reinforcement consisting
of 6-mm (0.24-in.)-diameter hoops at 265-mm (10.4-in.) cen-
ters. During quasi-static cyclic lateral loading tests, which
simulated seismic loading, available displacement ductility
factors of approximately 2 were found in these column units.
Evidence of bond degradation between the plain round lon-
gitudinal bars and the surrounding concrete was also ob-
served, which resulted in a softening of the measured lateral
load-versus-displacement relationships.
2. The two previously tested and damaged as-built column
units were retrofitted by adding reinforced concrete jackets.
In addition, two further as-built column units were retrofitted
by adding reinforced concrete jackets without being first sub-
jected to simulated seismic loading. The jackets consisted of
100-mm (3.94-in.)-thick new concrete containing new longi-
tudinal and transverse reinforcement. The surface of the as-
built columns had been lightly roughened by chipping before
the jackets were placed. The new longitudinal reinforcement
was placed either bundled in the four comers of the jacket
with single square hoops as transverse reinforcement, or dis-
tributed around the jacket with square and octagonal hoops
as transverse reinforcement. In the first case, the bundles of
longitudinal bars were 438 mm (17.2 in.) apart, and in the
second case the longitudinal bars were no more than 198 mm
(7.8 in.) apart. Results of the simulated seismic load tests
showed that the strength and stiffness of the jacketed columns
were up to 3 times those of the as-built columns. During
quasi-static cyclic lateral loading tests, with imposed nominal
displacement ductility factors of up to 6, very good energy
dissipation and only a small reduction in strength was ob-
served. These tests also showed that the effect of previous
damage to the as-built columns, and the two different rein-
158
(b) Unit SS4
forcing details used in the jacketed columns, had no signifi-
cant influence on the overall seismic performance of the jack-
eted columns.
3. The results of this investigation indicate that jacketing
with new reinforced concrete significantly improves the stiff-
ness, strength, and ductility of typical reinforced concrete
columns constructed according to early seismic codes. How-
ever, as was found in the investigation, this technique of retro-
fitting is labor-intensive.
ACKNOWLEDGMENTS
The financial assistance of the Earthquake and War Damage Comission
of New Zealand and the University of Canterbury is gratefully acknowl-
edged. Thanks are due to technicians P. Murphy and P. Coursey for assis-
tance with the construction and testing of the column units.
NOTATION
Ash total effective area of transverse confining reinforcement in direc-
tion of column under consideration
A8 gross area of column section
db bar diameter
/c' compressive cylinder strength of concrete
fy yield strength of longitudinal reinforcing steel
fyh yield strength of transverse reinforcing steel
he column depth
h" width of core of column measured to outside of the peripheral hoop
H lateralload
HAc! lateral load associated with the theoretical nominal flexural strength
of column calculated using the ACI 318 method and assuming a
strength reduction factor «!> = 1
£ shear span of column
P compressive load on column
Sh center-to-center spacing of hoop sets
~ horizontal displacement
~Y yield displacement
Pr ratio of area of longitudinal steel to gross area of column
ACI Structural Journal I March-April 1994
Ash, ACI = Ash in potential plastic hinge regions of columns recommended by
ACI 318-89
e rotation of column stubs
~ 1:~8
J.l.n nominal displacement ductility factor = 111fly
J.l.r real displacement ductility factor= (fl + 8£)/fly
LJ.l. cumulative nominal displacement ductility factor
REFERENCES
1. "Repair and Strengthening of Reinforced Concrete, Stone and Brick-
Masonry Buildings," Building Construction under Seismic Conditions in the
Balkan Regions, UNDP/UNIDO PROJECT RER/79/015, United Nations
Industrial Development Organization, Vienna, 1983, p. 231.
2. Endo T. et al., "Practices of Seismic Retrofit of Existing Concrete Struc-
tures in Japan," Proceedings of the Eighth World Conference of Earthquake
Engineering, V. 1, San Francisco, 1984, pp. 469-476. · Feb. 1987, pp. 61-76.
3. Aguilar, J.; Juarez H., Ortega, R.; and Iglesias, J., "Mexico Earthquake
of September 19, 1985. Statistics of Damage and of Retrofitting Techniques
in Reinforced Concrete Buildings Affected by the 1985 Earthquake," Earth-
quake Spectra, V. 5, No. 1, Feb. 1989, pp. 145-151.
4. Sugano, S., "Seismic Strengthening of Existing Reinforced Concrete
Buildings in Japan," Bulletin of the New Zealand National Society for Earth-
quake Engineering, V. 14, No.4, Dec. 1981, pp. 209-222.
5. Jara, M.; Hernandez, C.; Garcia, R.; and Robles, F., "Mexico Earthquake
of September 19, 1985. Typical Cases of Repair and Strengthening of Con-
crete Buildings," Earthquake Spectra, V. 5, No.1, Feb. 1989, pp. 175-193.
6. Brunsdon, D. R., and Priestley, M. J. N., "Assessment of Seismic Per-
formance Characteristics of Reinforced Concrete Buildings Constructed be-
tween 1936 and 1975," Bulletin of the New Zealand National Society for
Earthquake Engineering, V. 17, No.3, Sept. 1984, pp. 163-184.
7. Park, R., "Structural Concrete Developments in Buildings," Proceedings
of Silver Jubilee Conference, New Zealand Concrete Society, Wairakei, New
Zealand, Oct. 1989, pp. 9-25.
8. Hayashi, T.; Niwa, H.; and Fukuhara, M., "Strengthening Methods of the
ACI Structural Journal I March-April 1994
Existing Reinforced Concrete Buildings," Proceedings of the Seventh World
Conference on Earthquake Engineering, V. 4, Istanbul, 1980, pp. 89-97.
9. Bett, B. J.; Klingner, R. and Jirsa, J., "Lateral Load Response of Strength-
ened and Repaired Reinforced Concrete Columns," ACI Structural Journal, V.
85, No. 5, Sept.-Oct. 1988, pp. 499-508.
10. Rodriguez, M., and Park, R., "Repair and Strengthening of Reinforced
Concrete Buildings for Earthquake Resistance," Earthquake Spectra, V. 7, No.
3, Aug. 1991, pp. 439-459.
11. ACI Committee 318, "Building Code Requirements for Reinforced Con-
crete and Commentary (ACI 318-89/ACI 318R-89)," American Concrete Insti-
tute, Detroit, 1989, 353 pp.
12. "Code of Practice for the Design of Concrete Structures (NZS3101: 1982),"
Standards Association of New Zealand, Wellington, 1982, 127 pp.
13. Park, R., "Evaluation of Ductility of Structures and Structural Assem-
blages from Laboratory Testing," Bulletin of the New Zealand National Society
for Earthquake Engineering, V. 22, No.3, Sept. 1989, pp. 155-166.
14. Priestley, M. J. N., and Park, R., "Strength and Ductility of Concrete Bridge
Columns under Seismic Loading," ACI Structural Journal, V. 84, No. 1, Jan.-
15. Soesianawati, M. T.; Park, R.; and Priestley, M. J. N., "Flexural Ductility
of Reinforced Concrete Columns with Low Axial Load and Limited Transverse
Reinforcement," Proceedings ofPacific Conference on Earthquake Engineering,
V. 1, Wairakei, Aug. 1987, pp. 201-212.
16. Alcocer, S., and Jirsa, J. "Assessment of the Response of Reinforced Con-
crete Frame Connections Redesigned by Jacketing," Proceedings of the Fourth
U.S. National Conference on Earthquake Engineering, V. 3, May 1990, pp. 295-
304.
17. Tanaka, H., and Park, R., "Effectiveness of Transverse Reinforcement
with Alternative Anchorage Details in Reinforced Concrete Columns," Pro-
ceedings of Pacific Conference on Earthquake Engineering, V. 1, Wairakei,
Aug. 1987, pp. 225-235.
18. Rodriguez, M., and Park, R., "Seismic Load Tests on Reinforced Con-
crete Columns Strengthened by Jacketing, Research Report 92-9, Department
of Civil Engineering, University of Canterbury, Nov. 1992, p. 37.
19. Park, R., "Detailing of Transverse Reinforcement in Concrete Columns
for Ductility," Proceedings of the 4th United States National Conference on
Earthquake Engineering, V. 2, Palm Springs, May 1990, pp. 1037-1046.
159